Photocatalytic Degradation of Organic Dye by Sol-Gel-Derived Gallium-Doped Anatase Titanium Oxide Nanoparticles for Environmental Remediation

Abstract

Photocatalytic degradation of toxic organic chemicals is considered to be the most efficient green method for surface water treatment. We have reported the sol-gel synthesis of Gadoped anatase TiO2 nanoparticles and the photocatalytic oxidation of organic dye into nontoxic inorganic products under UV irradiation. Photodegradation experiments show very good photocatalytic activity of Ga-doped TiO2 nanoparticles with almost 90% degradation efficiency within 3 hrs of UV irradiation, which is faster than the undoped samples. Doping levels created within the bandgap of TiO2 act as trapping centers to suppress the photogenerated electron-hole recombination for proper and timely utilization of charge carriers for the generation of strong oxidizing radicals to degrade the organic dye. Photocatalytic degradation is found to follow the pseudo-first-order kinetics with the apparent 1st-order rate constant around 1.3×[superscript]10-2[/superscript] min-1. The cost-effective, sol-gel-derived TiO2 : Ga nanoparticles can be used efficiently for light-assisted oxidation of toxic organic molecules in the surface water for environmental remediation.

(ProQuest: ... denotes non-US-ASCII text omitted.)

Arghya Narayan Banerjee 1 and Sang Woo Joo 1 and Bong-Ki Min 2

Recommended by Vo-Van Truong

1, School of Mechanical Engineering, Yeungnam University, Gyeongsan 712-749, Republic of Korea
2, Center for Research Facilities, Yeungnam University, Gyongsan 712-749, Republic of Korea

Received 20 January 2012; Revised 12 March 2012; Accepted 13 March 2012

1. Introduction

Titanium dioxide (TiO₂ ) is one of the most important wide bandgap metal oxides which is having a vast range of applications from paint to sunscreen to food coloring to photocatalyst, hydrogen production, storage medium, sensors, solar cells, organic waste management, and various biological and health-related applications [1-13]. Because of its wide range of properties, TiO₂ bulk films as well as nanostructured materials become the subject of intense research within the global scientific community. In general, TiO₂ has two stable crystalline structures: anatase and rutile [14]. Rutile is preferred to anatase for optical applications because of its higher refractive index, whereas anatase is preferred for all the applications related to photocatalytic activity, gas sensing, and solar cells, due to its higher mobility and catalytic properties [15, 16].

Amongst various TiO₂ nanostructures, titania nanoparticles have specific advantages in the enhancement of light absorption due to the large fraction of surface atoms. Interband electron transition is the primary mechanism of light absorption in pure semiconductors. These transitions are direct as the momentum gain by the electron from light wave is small in comparison with πh/a ("a " is the lattice constant). This absorption is small in direct-forbidden gap semiconductors, as in the case for TiO₂ , for which the direct electron transitions between the band centers are prohibited by the crystal symmetry. However, momentum is not conserved if the absorption takes place at the boundary of the crystal, for example, at the surface or at the interface between two crystals, which leads to the indirect electron transitions that can result in the essential enhancement of light absorption. This means that considerable enhancement of the absorption can be observed in small nanocrystals where the surface-to-volume ratio is very high and the fraction of the surface atoms is sufficiently large. The particle size at which the interface enhancement of the absorption becomes significant is around 20 nm or less. An additional advantage obtained in nanoparticles in the few nanometer size regimes is that the large surface-to-volume ratio makes possible the timely utilization of photogenerated carriers in interfacial processes [1, 17]. Additionally, the doping of TiO₂ nanoparticles is performed for improved photocatalytic activities by reducing the band gap of TiO₂ to utilize the wider fraction of solar radiation, especially the visible and near infrared (NIR) parts [4, 18-20]. Many efforts have been expended to narrow the TiO₂ band gap by substitutional doping. According to the crystal structure of TiO₂ , it appears that replacement of Ti⁴⁺ with any cation is relatively easier than to substitute O^2- with any other anion due to the difference in the charge states and ionic radii. Cationic doping of TiO₂ with transition and rare earth metals has been extensively studied [18-24]. While several authors have reported that transition metal ion doping decreases the photothreshold energy of TiO₂ , there is also an increase in thermal instability and a decrease in carrier lifetimes [25], which limits overall conversion efficiencies. Therefore, it is clear that there is always scope for improvement in the photoactivity of TiO₂ nanostructures either by applying various new dopants or (and) adopting different doping parameters through different deposition processes and conditions. Amongst various photocatalytic applications of TiO₂ , photocatalytic degradation of toxic organic chemicals (especially organic dyes generated as industrial wastes and released in the surface water without proper treatment) is considered to be the most efficient green method for organic waste management in terms of photosensitized TiO₂ -assisted oxidation of organic pollutants for surface water treatment, recovery of precious metals via TiO₂ -assisted reduction, organic synthesis, photokilling activity, and self-cleaning activity among others [26-35].

As far as the syntheses of undoped and doped TiO₂ nanostructures are concerned, both solution-based chemical techniques as well as vacuum-based physical techniques [1, 7, 9] have been adopted. In the current study, we have reported the sol-gel syntheses and characterizations of Ga-doped anatase TiO₂ nanoparticles (TiO₂ : Ga) and investigated the photocatalytic oxidation of organic dyes for environmental remediation. Apparently, sol-gel deposition process is preferred (at least in the research scale) over vacuum-based as well as hydrothermal syntheses, mainly because of its simplicity and cost-effectiveness in terms of materials, design, process, and implementation. As far as the reason behind the adoption of Ga as the doping material is concerned, previously few authors reported the improved photocatalytic activities of TiO₂ : Ga (and TiO₂ : Ga/I co-doped) nanomaterials in terms of solar hydrogen production and photodecomposition of various organic compounds used in disinfectants and pesticides [36-41]. The comparable ionic radii of Ta⁺⁴ (68 pm) and Ga⁺³ (62 pm) appears to be favorable for the substitutional doping of Ga atoms into the Ti matrix to improve the optical, electronic, and photocatalytic properties. To best of our knowledge, no report has been published on the photocatalytic oxidation of organic dyes by anatase TiO₂ : Ga nanoparticles for organic waste treatments, which is one of the burning issues in the industrial waste management and, hence, needs much attention.

As far as the difference between our work against the previously reported sol-gel-derived Ga-doped (or Ga/I co-doped) TiO₂ nanomaterials is concerned [37, 38], both of these reports indicated the formation of phase impure TiO₂ nanopowders with a mixture of anatase and rutile (as well as a small fraction of brookite) phases under Ga (or Ga/I) (co-)doping, whereas we have been able to obtain phase pure anatase TiO₂ nanoparticles (as shown later) via our sol-gel route. It has been reported previously that phase pure anatase TiO₂ (along with commercially available Degussa P-25) has the highest hydroxyl radical ( OH * ) formation rate under photoirradiation, which is considered to be the main species responsible for the photooxidation of organic compounds [10]. Therefore, it is desirable to synthesize phase-pure anatase TiO₂ for efficient photodegradation of organic molecules, and, in that respect, our process to synthesize phase-pure Ga-doped anatase TiO₂ nanoparticles may become very important for efficient photocatalytic organic waste management. Also, for the application of waste water treatment via TiO₂ : Ga-based photocatalytic degradation of organic molecules, photocatalytic oxidation of water soluble organic dyes (like Rhodamine B, as in our case) should be preferred as the testing agent against sparingly soluble organic substances (like 2-chlorophenol, 1,4-dichlorobenzene, and so forth, as used in [37, 38]), because the water soluble toxic dyes released from textile industries in the surface water as waste products pose considerable threats to the environment and, hence, warrants proper attention.

2. Experimental

2.1. Materials

Undoped and Ga-doped anatase TiO₂ nanopowders were synthesized by the sol-gel route using titanium isopropoxide [Ti_{{OCH(CH3 )2 }4} ] and gallium(III) oxide (Ga₂ O₃ ), (all from Sigma Aldrich, Korea, 99.9% pure) as Ti and Ga sources, respectively. Isopropanol, acetic acid (99.0%, Duksan Pure Chemicals, Republic of Korea), and deionized (DI) water is used for precursor solution preparation. For photocatalytic experiment, Rhodamine B ( ^Rh B , Sigma Aldrich, Republic of Korea) is used as the organic dye in the aqueous solution.

2.2. Method

Firstly, 0.05 mol of titanium isopropoxide is dissolved into isopropanol with a 1 : 10 molar ratio, respectively. In order to stabilize the Ti solution, acetic acid is then added to the solution until the pH of the solution is in between 3.0 and 3.5. The resulting solution is stirred for 60 min at room temperature until a homogeneous sol is prepared. Then, the sol is aged for 24 hrs at ambient condition to get the gel. Thereafter, the gel is air-dried at 100°C to remove the solvents. The dried gels are then calcined in air ambient at 700°C (in a furnace, Thermo Fisher Scientific, Republic of Korea) for 2½ hours to get the nanocrystalline powder which is then grounded in an agate mortar to get TiO₂ nanoparticles. For the syntheses of Ga-doped TiO₂ , 1.0 at.% Ga doping are performed in the nominal solution by dissolving stoichiometric amounts of Ga₂ O₃ in 0.4 mol isopropanol and deionized water (with 10 : 1 molar ratio) and stirred for 30 min at room temperature. This doping solution is then added to the Ti sol prepared by above-mentioned method and stirred for 60 min at room temperature followed by doped sol preparation and corresponding gelation as mentioned above. The layout of the deposition process is furnished in Figure 1.

Figure 1: Flow chart of the sol-gel syntheses of undoped and Ga-doped TiO₂ nanoparticles. Two different colors are used to separate the doping processes from the nondoping case.

[figure omitted; refer to PDF]

The probable chemical reactions taking place at different stages of the sol-gel process are furnished as sources. [figure omitted; refer to PDF] [figure omitted; refer to PDF] [figure omitted; refer to PDF] [figure omitted; refer to PDF] [figure omitted; refer to PDF]

Titanium isopropoxide (Ti-OR) is reacted with isopropyl alcohol (^Pri -OH) to form alkoxide which is then hydrolyzed to form Ti-OH, followed by condensation to get Ti-O-Ti group along with the by-products of water and alcohol. This produces the sol, which is then aged to get long chains of ... Ti-O-Ti-O-Ti-O- ... , terminated by hydroxyl groups, resulting in the thickening of the sol to form the gel. Finally, the gel is dried and calcined in oxygen (or air) atmosphere to get TiO₂ .

2.3. Characterizations

The structural properties of the as-synthesized undoped and Ga-doped TiO₂ nanoparticles are studied by X-ray diffractometer (XRD, Philips XPertPowder) and Raman spectroscopy (Confocal Raman Microscope, Model number alpha300R, WITec, Germany) with excitation of 532 nm. Microstructural analysis of the nanoparticles is performed by scanning transmission electron microscope (STEM, Tecnai G² F20). For high-resolution (HR) TEM characterizations, the nanoparticles are dispersed in alcohol, followed by stirring and drop casting onto the C-coated Cu grid and drying. Optical diffuse reflectance measurements are performed by using a UV-Vis-NIR spectrophotometer (Shimadzu UV-160A) in the spectral range of 300 to 2500 nm. X-ray photoelectron spectroscopic (XPS, K-Alpha, Thermo Fisher Scientific, Republic of Korea) measurements are done by using X-ray Al _Kα source. The wide range survey spectra are carried out within the binding energy (B.E.) range of 0.0 to 1350.0 eV with a pass energy of 200.0 eV, whereas the high resolution spectra for O 1s, Ti 2p, Ga 3d electronic structures are measured with a 50.0 eV pass energy.

2.4. Photocatalytic Characterization

The photocatalytic activity of Ga-doped TiO₂ is performed by recording the degradation of ^Rh B in aqueous solution under UV radiation. The home-built photoreactor consists of two 16 W UV tube (Fisher Scientific, Republic of Korea) using 254 nm wavelength placed 8 cm away from the reaction vessel containing the dye solution. Before reaction, the dye solution (50 mL) is prepared by mixing ^Rh B and DI water in a molar ratio of 3.6 × 10^-6 : 1 and stirred by a magnetic stirrer (Fisher Scientific, Republic of Korea) for 45 min in dark condition to ensure the proper adsorption-desorption equilibrium of dye molecules on the nanoparticle surfaces to act as efficient photocatalysts. The organic dye degradation experiment is performed at different times to observe the decrement in the dye concentration by taking digital photographs of the solution until the whole solution is decolored to ensure complete degradation of the dye molecules. To determine the dye concentration within the degraded solution, UV-Vis spectrophotometric measurements (Shimadzu UV-160A) of the dye solution oxidized at different times are performed within the spectral range of 300 to 800 nm. Before the optical measurements, the dye is separated from the nanoparticles by centrifuging the solution for 10 min (Model 1-14, Sigma, Germany).

3. Results and Discussion

3.1. XRD and HRTEM Analyses

XRD analyses of undoped and Ga-doped TiO₂ nanoparticles are shown in Figure 2(a). The peaks show the formation of anatase TiO₂ with I4₁ /amd space group (1999 JCPDS ICCD File Card nymber 83-2243). No peaks of initial constituents and any by-product have been observed for our samples to confirm the formation of phase pure anatase TiO₂ nanomaterials along with the proper doping within the Ga-doped sample. It has been observed that the peak intensities of both the patterns are almost similar, whereas the full width at half maxima (FWHM) of the peaks of undoped sample are slightly lesser than that of doped one. Especially, the broad peak between 36° to 40° is found to be fairly resolved in the undoped sample (as shown in Figure 2(b) through deconvolution) into three distinct peaks ((103), (004), and (112)), whereas the doped sample shows a broad hump around 37.8°. Similarly, the well-resolved (105)/(211) and (116)/(220) peaks in the undoped sample tend to merge in the doped sample. Also a relative comparison of the (101) peaks of as-synthesized undoped and metal-doped samples (cf. Figure 2(c)) reveals that the FWHM increases in the doped sample against undoped samples, with the peak positions remaining almost identical (and this trend is valid for all other peaks of the samples as well). It is well-known that "particle-broadening effect" generally increases the FWHMs of the diffraction peaks in nanoparticles [42]. But the HRTEM imaging reveals (shown later) that the particle sizes for both undoped and doped samples are almost similar, indicating negligible size effect within the samples. Therefore, it is speculated that "strain-broadening" is highly effective in the as-synthesized doped samples for the relative increment in the FHHMs of the XRD peaks. Generally, in a polycrystalline aggregate, the individual crystallites in a grain is not free to deform in the same way as an isolated crystal would, if subjected to the same deforming force. As a result of this restraint by its neighbors, a grain in a polycrystalline aggregate is usually in a state of tension or compression, which produces uniform or nonuniform strain within the sample and reflected either by a shift in the diffraction peaks (for uniform strain) or by a peak broadening (also called "strain-broadening" for nonuniform strain) without changing the peak position [42]. As mentioned earlier, the ionic radius of Ti⁺⁴ is 68 pm, whereas the same for Ga⁺³ is 62 pm. Apparently, the doping of Ga⁺³ into lattice Ti⁺⁴ sites (substitutional doping) as well as in the interstitial sites creates nonuniform strain within the doped samples which is reflected in the peak-broadening effect of the XRD peaks of TiO₂ : Ga nanopowders. To quantitatively determine the effective crystallite size, ^Lcrysteff and strain, ^{[straight epsilon]eff} within the undoped and metal-doped samples, the FWHMs ( β , measured in radian) of the XRD peaks for different 2 θ values are expressed as [figure omitted; refer to PDF] where λ is the incident X-ray wavelength (1.5406 Å). A plot of βcos θ/λ versus sin θ/λ will be a straight line, slope of which will give the estimation of the effective strain, whereas the intercept on βcos θ/λ axis will carry the information of the effective crystallite size. Figure 2(d) shows the size-strain analysis plots of undoped and Ga-doped TiO₂ samples and reveal the average crystallite sizes around 10.0 to 12.0 nm, whereas the strain increases from 7.53 × 10^-3 (for undoped samples) to a maximum of 2.71 × 10^-2 (for Ga-doped samples), which is consistent with the above discussions regarding the increasing trend of FWHMs between undoped and doped samples. A comparison between size-strain values for two types of samples is furnished in Table 1.

Table 1: Different structural parameters of undoped and metal-doped TiO₂ .

Parameters	TiO2	TiO2 : Ga
Effective crystallite size (Lcrysteff ) (nm) (from XRD data)	10.4	11.8
Effective strain ([straight epsilon]eff ) (from XRD data)	7.53 × 10-3	2.71 × 10-2
Average particle size (Lavg ) (nm) (from HRTEM micrographs)	15.3	17.8
Atomic ratio (Ti : O : Ga) (from EDX data)	33.87±0.51 : 66.12±0.65	33.49±0.52 : 65.51±0.66 : 0.99±0.047

Figure 2: (a) XRD patterns of undoped and Ga-doped TiO₂ nanoparticles. (b) Magnified XRD graph of the undoped TiO₂ sample between 36° and 40° of 2 θ values, showing the resolution of (103), (004), and (112) peaks after deconvolution. (c) Magnified (101) peaks of the undoped and Ga-doped TiO₂ nanoparticles showing the strain broadening effect within the doped samples. (d) Size-strain analyses of undoped and Ga-doped TiO₂ nanoparticles.

[figure omitted; refer to PDF]

Figures 3(a) to 3(d) show the HRTEM micrographs of undoped and Ga-doped TiO₂ nanoparticles for two different magnifications. High-mag images show the proper lattice spacing of TiO₂ . The insets of Figures 3(a) and 3(c) represent the corresponding size distributions of as-synthesized nanoparticles. The sizes are calculated from the image analysis of the micrographs assuming the particle area to be the projection of spherical particles. The average particle sizes (^Lavg ) are obtained as 15.3 nm (undoped TiO₂ ) and 17.8 nm (TiO₂ : Ga), respectively, and consistent with the crystallite size calculations from XRD data (Figure 2(d)) for polycrystalline nanoparticles. Also the insets of Figures 3(b) and 3(d) represent the selected area electron diffraction (SAED) patterns of the samples indicating the characteristics circles originating from the corresponding crystals planes of anatase TiO₂ . Figure 4 shows the EDX data of undoped and metal-doped TiO₂ nanoparticles indicating the presence of elemental constituents within the as-synthesized samples. The atomic percentages of Ti, O, and Ga are furnished in Table 1.

HRTEM micrographs of (a)/(b) undoped TiO₂ and (c)/(d) TiO₂ : Ga at two different magnifications. Insets of (a)/(c) are the corresponding size distributions of nanoparticles. Insets of (b)/(d) are the SAED patterns of undoped and Ga-doped TiO₂ nanoparticles, respectively.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

(c) [figure omitted; refer to PDF]

(d) [figure omitted; refer to PDF]

EDX analyses of (a) undoped and (b) Ga-doped TiO₂ nanoparticles, indicating the presence of elemental Ti, O, and Ga.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

3.2. XPS Analyses

The chemical and electronic structures of as-synthesized undoped and metal-doped samples are determined by XPS analysis and shown in Figure 5. Figure 5(a) shows the high-resolution XPS spectra of undoped and metal-doped TiO₂ nanoparticles of O 1s region, which, after deconvolution, can principally be assigned to lattice oxygen (Ti ^x -O: x=+3,+4 ) and surface-adsorbed hydroxyl groups (Ti-OH). Table 2 lists the results of curve fitting of XPS spectra for undoped and metal-doped TiO₂ . As seen in the table, the Ti-O peaks are formed at a binding energy (B.E.) around 530.5 eV for all the samples, which is slightly blue-shifted (+0.7 eV) from the standard value of ~529.8 eV [40]. Similar kinds of blue-shifts are reported by others [43, 44] as well and attributed to the presence of larger number of surface hydroxyl groups within the samples. In our case also, the comparatively higher percentage of Ti-OH concentration against previously reported values (8.0-25.0%) [40, 45] clearly depicts the higher OH^- -terminated surface in our samples and may be attributed to the higher surface-to-volume ratio in our nanoparticles that created larger number of surface adsorption sites for hydroxyl ions. Also the relative concentrations of Ti-OH peaks are found to increase in the Ga-doped sample against undoped one, suggesting that the Ga-doped TiO₂ enhances the adsorption of OH^- ions, which consequently increases the hydrophilicity of the nanoparticles [43]. As far as the photocatalytic activity is concerned, it has been reported that the so-called surface "titanol" groups (Ti-OH) in TiO₂ photocatalysts act as the hole scavenger to form the main oxidizing site to improve the photoactivity of TiO₂ [44]. Therefore, Ga-doped TiO₂ with higher Ti-OH concentration is expected to produce larger number of hydroxyl groups to enhance its photocatalytic activity during photolysis. Also, the substitution of a Ga⁺³ into Ti⁺⁴ sites will create oxygen vacancy, which serves as an electron trap and effectively enhances the lifetime of the hole, which, in turn, helps in the oxidative degradation of pollutants. As far as the presence of Ti⁺⁴ states in the as-synthesized samples is concerned, Figure 5(b) shows the typical high-resolution spectra of undoped and Ga-doped TiO₂ nanoparticles, indicating the spin-orbital splitting of Ti 2p_3/2 and Ti 2p_1/2 photoelectrons in anatase TiO₂ at B.E. values around 459.3 eV and 465 eV, respectively. The FWHMs of both the peaks are found to increase in the doped sample against the undoped one (cf. Table 2), which implies the presence of less-oxidized metals in the doped samples, indicating the existence of Ti⁺³ species within the TiO₂ matrix. A closer look into the Ti⁺⁴ 2p_3/2 peaks of doped sample reveals the existence of a shoulder around 457.8 eV (cf. Figure 5(c) and Table 2), which originates from Ti⁺³ 2p_3/2 state of photoelectrons [43]. The presence of this lesser oxidation states of Ti ions in the doped samples indicates the formation of higher percentage of loosely bound Ti⁺³ -O bonds against stronger Ti⁺⁴ -O bonds, which enhances the photoelectron transfer process between anions and cations and, thus improves the photocatalytic activities of the doped nanoparticles [44]. For Ga-doped samples, Ga 3d spectrum is shown in Figure 5(d), which after deconvolution, reveals the presence of two peaks, the lower B.E. peak (at 21.3 eV) originates from elemental Ga in Ga₂ O₃ and the higher one (at 22.7 eV) is speculated to be due to Ga-Ti complex within the TiO₂ : Ga nanoparticles. Similar unknown Ga 3d peak (at 22.8 eV B.E.) is observed by Ghita et al. [46] in their Ti-oxidized GaAs samples and speculated to be originated from Ga-Ti complex. The relative area of these two peaks (cf. Table 2) shows the lesser percentage of Ga₂ O₃ formation against Ga-Ti complex and may be attributed to the greater fraction of interstitial doping with respect to substitutional doping in our highly doped TiO₂ : Ga nanoparticles. This argument finds support from the EDX data (cf. Table 1), where it is found that the doped sample becomes more metal excess against the undoped sample. Ideally, the chemical formula for metal-doped TiO₂ would be _Ti1-xO2My ; (M is dopant ion, x is substitution of Ti by M ions, and y is doping concentration). For 100% substitutional doping of M ions in the Ti sites, x will assume the value of y . Else x<y , when interstitial doping is also present, which is thermodynamically more practical situation. Therefore, the presence of excess nonstoichiometric metal content in our Ga-doped samples indicates a larger fraction of interstitial doping within the TiO₂ matrix with respect to substitutional doping. It should also be mentioned in this connection that, in the XPS survey spectra of undoped and metal-doped samples (not shown here), presence of C 1s peaks is also observed, indicating the existence of small C contents within our samples, which may be attributed to the adventitious hydrocarbon from specimen handling and (or) pumping oil of the XPS system.

Table 2: Various XPS data of O and Ti species within the undoped and metal-doped TiO₂ nanoparticles.

Electronic levels	Fitting parameters	TiO2	TiO2 : Ga
O 1s	Ti-O	B.E. (eV)	530.5	530.5
FWHM (eV)	1.02	1.06
Area1 (%)	66.0	58.4
Ti-OH	B.E. (eV)	531.7	531.9
FWHM (eV)	2.77	3.34
Area1 (%)	34.0	41.6
Ti 2p	Ti+4 2p3/2	B.E. (eV)	459.3	459.3
FWHM (eV)	1.02	1.08
Area2 (%)	69.3	69.1
Ti+4 2p1/2	B.E. (eV)	465.0	465.0
FWHM (eV)	1.68	1.75
Area2 (%)	30.7	30.9
Ti+3 2p3/2	B.E. (eV)	--	457.8
FWHM (eV)	--	1.02
Area3 (%)	--	6.1
Ga 3d	Due to elemental Ga in Ga+3 -O	B.E. (eV)	--	21.3
FWHM (eV)	--	5.27
Area4	--	33.3
Due to formation of Ga-Ti complex	B.E. (eV)	--	22.7
FWHM (eV)	--	2.74
Area4	--	66.7

¹ With respect to the entire area under the O 1s curve.

² With respect to the entire area under the Ti 2p curve.

³ With respect to the area under the Ti 2p_3/2 curve only.

⁴ With respect to the entire area under the Ga 3d curve(s).

XPS spectra of (a) O 1s, (b) Ti 2p, (d) Ga 3d photoelectrons in undoped and Ga-doped TiO₂ nanoparticles. (c) Deconvolution of Ti 2p peaks of doped samples to show the existence of Ti⁺⁴ and Ti⁺³ ions.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

(c) [figure omitted; refer to PDF]

(d) [figure omitted; refer to PDF]

3.3. Optical Characterizations

Raman spectra of undoped and Ga-doped TiO₂ nanoparticles are shown in Figures 6(a) and 6(b), respectively. The peaks placed at 145, 398, 517, and 639 cm^-1 (indexed as "A") for both the spectra are typical characteristics of the anatase phase of TiO₂ , thus corroborating the XRD data [37]. A small peak at 318 cm^-1 is observed in the doped sample and speculated to be originating from vibrations of Ga₂ O₃ , as observed previously by others as well for β -Ga₂ O₃ [47, 48], thus supporting the XPS data of our doped samples showing the presence of Ga₂ O₃ within the TiO₂ matrix. The spectral variations of diffuse reflectance (R) of undoped and Ga-doped TiO₂ nanoparticles are shown in Figure 6(c). For undoped and doped samples, the strong absorption edge, which corresponds to the near-band-edge emission, is found to be less than 400 nm and typical of anatase TiO₂ . But a relative comparison of the visible parts of the reflectance spectra of both the samples (cf. Inset-(i) of Figure 6(c)) depicts slight absorption into the visible region for the doped samples, owing to the charge-transfer between the Ga d electrons and the TiO₂ conduction band, as shown schematically in Figure 6(d). Apparently, metal doping creates a defect level within the bandgap of TiO₂ and electrons can be excited from the defect state to the TiO₂ conduction band by photons with energy (h_νD ) less than the bandgap energy (h_νG ) , thus expanding the active region of solar spectrum for photocatalytic activity of TiO₂ into the visible range. Additionally, the metal doping creates trapping centers for electrons to decrease the electron-hole recombination during irradiation to enhance the photoactivity [4]. To determine the exact bandgap values of two types of nanoparticles, an approximate relation between R , absorption coefficient ( α ), and scattering coefficient (s ), based on Kubelka-Munk theory, is adopted, assuming no transmission within the powdered samples of few millimeters thickness [49], [figure omitted; refer to PDF] Although both α and s are functions of the energy of incident radiation (hν) , but at the region of strong absorption, s is far less wavelength dependent than α and, hence, can be considered as constant at the onset of absorption edge. To determine the bandgap values, a graph between ^(αhν)2 and hν is plotted at and around the fundamental absorption region (cf. inset-(ii) of Figure 6(c)), which clearly shows that the fundamental bandgaps of both the samples are identical and fall at 3.32 eV. This indicates no size effect in our undoped and Ga-doped TiO₂ nanoparticles, as depicted from the HRTEM analysis.

Raman spectra of (a) undoped, (b) Ga-doped TiO₂ nanoparticles. Peaks indexed as "A" are originating from vibrations associated with anatase TiO₂ , whereas the peak indexed as "B" in the doped sample is from Ga₂ O₃ . (c) UV-Vis-NIR diffuse reflectance spectra of undoped and Ga-doped TiO₂ nanoparticles. Inset (i) represents the visible part of reflectance spectrum, and inset (ii) represents the direct bandgap calculations of undoped and Ga-doped TiO₂ nanoparticles, (d) approximate energy level diagram of doped TiO₂ showing the defect level generation for improved photoactivity and the photocatalytic degradation mechanism (the diagram is not to the scale).

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

(c) [figure omitted; refer to PDF]

(d) [figure omitted; refer to PDF]

3.4. Photocatalytic Activity

The role of Ga-doped TiO₂ nanoparticles as efficient photocatalyst is determined by the photocatalytic degradation of ^Rh B under UV light at different irradiation times, and corresponding time effluence of the absorption spectra of the degraded dye solution is shown in Figure 7(a). The characteristics absorption peak intensity of ^Rh B at 554.1 nm is shown to be significantly decreased within 2 hrs of irradiation, and the decoloration of the dye is completed within 3 hrs, as shown by the digital photographs of the dye solutions in the inset of Figure 7(a). The overall degradation time is lesser than that of previously reported degradation results of ^Rh B through TiO₂ photocatalyst [50], indicating higher efficiency of degradation of our Ga-doped TiO₂ nanoparticles. Also, a slight blue-shift of the absorption peak with an increase in time indicates the formation of some N-de-ethylated intermediates during photodegradation of ^Rh B , as reported by others as well [51]. The entire photocatalytic process for organic dye degradation is furnished as follows [3, 4] [figure omitted; refer to PDF] [figure omitted; refer to PDF] [figure omitted; refer to PDF] [figure omitted; refer to PDF] The incident photons, having energy (h_νin ) greater than the threshold (h_νD ) , generate electron-hole pairs at the nanoparticle surface (7), a large part of which takes part in the photocatalytic process because of the suppression of ^e- -^h+ recombination by charge trapping of the defect levels generated by Ga doping. Depending upon the exact conditions, the holes at the valence band and electrons at the conduction band react with water and adsorbed oxygen to produce hydroxyl radicals ( OH * ) and superoxide anion radicals ( ^{*_O2 -} ), respectively, through some other products like ^OH- , O * OH , _H2O2 , and O₂ (8), (9). Finally, these radicals with high redox oxidizing potential decompose the organic dye into CO₂ , _H2 O , and other inorganic nontoxic products (10). A schematic representation of the proposed mechanism is shown in Figure 6(d). Further, it has been reported previously [52, 53] that the OH * radical, which is considered to be one of the main species with high redox oxidizing potential to decompose the organic dye, can be detected by photoluminescence (PL) technique and quantitatively scaled to compare the photocatalytic performance of various photocatalysts. According to this methodology, phase pure anatase TiO₂ (along with Degussa P-25) showed very high formation rate of OH * , so also better photoactivity [52]. Therefore, a detailed PL analysis of Ga-doped TiO₂ as a function of doping concentration is necessary to correlate the OH * formation rate and the photocatalysis of our nanoparticles, which is further course of our research.

(a) Time effluence on photocatalytic degradation of Rhodamine B dye using TiO₂ : Ga nanoparticles as the photocatalyst under UV irradiation. Inset represents the digital photographs of dye solution at different irradiation times to show color change due to photocatalytic degradation of ^Rh B . (b) Pseudo-first-order kinetic plots and apparent rate constant of ^Rh B degradation.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

To quantitatively determine the performance of the photocatalyst, the percentage degradation (% d ) is calculated by measuring relative decrease in the dye concentration against the initial value by estimating the peak absorbance of the absorption spectra of the dye solution at different irradiation times according to the following relation: [figure omitted; refer to PDF] where ^Adye0 and ^Adyeτ are the peak absorbance of dye solution at initial concentration and after irradiation time τ , respectively. The degradation is found to be 50%, 80%, and 90% after irradiation times of 1 hr, 2 hr, and 3 hr, respectively. It is well known that the photocatalytic degradation kinetics follow the Langmuir-Hinshelwood model [54], which at low dye concentration can be approximated to an apparent 1st-order kinetic reaction according to the following relation [54]: [figure omitted; refer to PDF] where ^Cdye0 and ^Cdyeτ are the dye concentrations at initial and at irradiation time τ and ^kapp is the apparent 1st-order rate constant. A plot of ln (^Cdye0 /^Cdyeτ ) versus τ will be a straight line, slope of which will carry the information of ^kapp . Figure 7(b) represents the ln (^Cdye0 /^Cdyeτ ) versus τ plot for our samples, and the value of ^kapp is obtained as 1.3 × 10^-2 min^-1 , which is found to be better than the previously reported values [50] of similar initial ^Rh B concentrations, photocatalyzed with TiO₂ , indicating better photoactivity of our Ga-doped TiO₂ nanoparticles. Also, to compare the photocatalytic performance of our sample (1.0 at.% Ga-doped TiO₂ nanoparticles) with standard Degussa P-25 (as-received) samples, we have observed an 80% degradation of ^Rh B within 120 min of irradiation, which is better than the Degussa P-25 TiO₂ powders, where ^Rh B is found to be degraded upto 80-90% within 150 min [55]. Further, the results are found to be better for 2.0 at.% Ga-doped samples against Degussa P-25, where we have observed almost 90% degradation of ^Rh B within 90 min of irradiation.

As far as the effect of doping concentrations on the photocatalytic activity of the Ga-doped TiO₂ nanoparticles is concerned, we have observed a gradual increase in the photoactivity of the nanoparticles from 0.1 at.% to 2.0 at.% Ga-doped samples, beyond which the photoactivity started to decrease. It is speculated that, beyond the optimal doping concentration, the Ga⁺³ ions may not act favorably as the charge trapping centers to reduce the electron-hole recombination, rather become recombination centers through "quantum tunneling" to affect the photoactivity significantly, as observed previously for Fe⁺³ -doped TiO₂ nanomaterials [41, 56]. Similar effect is also observed for Ga-doped TiO₂ nanoparticles as well [36, 37], where the photoactivity is found to reach an optimum value for a certain dopant concentration (2 mol.% or 2-5 at%). In the current paper, we have reported all the data against 1.0 at% Ga doping, but a detailed study on the effect of doping concentration on the photocatalytic activity of Ga-TiO₂ for photodegradation of organic dye (such as ^Rh B , Methyl Orange, Orange-II, etc.) is underway and will be reported in due course.

4. Conclusions

Photocatalytic degradation of organic dye solution (Rhodamine B) is performed using Ga-doped TiO₂ nanoparticles as the photocatalyst. The nanoparticles are synthesized via sol-gel deposition process by using titanium isopropoxide and gallium oxide as the starting materials. XRD and Raman measurements confirm the pure anatase phase formation of the nanoparticles. Size-strain analysis shows that the doped samples are slightly strained due to the presence of Ga ions within TiO₂ matrix either as substitutional and (or) interstitial dopants. The nanoparticle size is found to be around 15 nm and confirmed via HRTEM imaging. XPS analysis shows the presence of Ti⁺³ and surface titanol groups, which enhance the photoactivity of the nanoparticles. Optical reflectance measurements showed a slight red-shift of the bandgap of the doped sample due to the formation of defect state within the bandgap of TiO₂ to improve the photocatalytic activity of TiO₂ . Degradation of dye solution is done under UV irradiation and found to be highly efficient. The apparent 1st-order rate constant is found to be around 1.3 × 10^-2 min^-1 , which is relatively higher than the previously reported values of TiO₂ -assisted photodegradation of ^Rh B with similar initial concentrations. The cost-effective, sol-gel-derived TiO₂ : Ga nanoparticles can be used efficiently for light-assisted oxidation of toxic organic molecules in the surface water for environmental remediation.

Acknowledgments

Financial support of the World Class University Grant no. R32-2008-000-20082-0 of the Korea Research Foundation is acknowledged. The authors are also grateful to Ji-Hye Park and Dong-Ju Lim of UNIST Central Research Facilities for Raman and XPS characterizations, respectively.

References

[1] A. N. Banerjee, "The design, fabrication, and photocatalytic utility of nanostructured semiconductors: focus on TiO₂ -based nanostructures," Nanotechnology, Science and Applications , vol. 4, no. 1, pp. 35-65, 2011.

[2] A. Fujishima, K. Honda, "Electrochemical photolysis of water at a semiconductor electrode," Nature , vol. 238, no. 5358, pp. 37-38, 1972.

[3] A. L. Linsebigler, G. Lu, J. T. Yates, "Photocatalysis on TiO₂ surfaces: principles, mechanisms, and selected results," Chemical Reviews , vol. 95, no. 3, pp. 735-758, 1995.

[4] A. Zaleska, "Doped-TiO₂ : a review," Recent Patents on Engineering , vol. 2, no. 3, pp. 157-164, 2008.

[5] K. Shankar, J. I. Basham, N. K. Allam, O. K. Varghese, G. K. Mor, X. Feng, M. Paulose, J. A. Seabold, K. S. Choi, C. A. Grimes, "Recent advances In the use of TiO₂ nanotube and nanowire arrays for oxidative photoelectrochemistry," Journal of Physical Chemistry C , vol. 113, no. 16, pp. 6327-6359, 2009.

[6] K. Juodkazis, J. Juodkazyte, E. Jelmakas, P. Kalinauskas, I. Valsiunas, P. Miecinskas, S. Juodkazis, "Photoelectrolysis of water: Solar hydrogen-achievements and perspectives," Optic Express , vol. 18, no. 102, pp. A147-A160, 2010.

[7] X. Chen, S. S. Mao, "Titanium dioxide nanomaterials: synthesis, properties, modifications and applications," Chemical Reviews , vol. 107, no. 7, pp. 2891-2959, 2007.

[8] A. Wold, "Photocatalytic properties of TiO₂ ," Chemistry of Materials , vol. 5, no. 3, pp. 280-283, 1993.

[9] U. Diebold, "The surface science of titanium dioxide," Surface Science Reports , vol. 48, no. 5-8, pp. 53-229, 2003.

[10] S. Liu, J. Yu, M. Jaroniec, "Anatase TiO₂ with dominant high-energy {001} facets: synthesis, properties, and applications," Chemistry of Materials , vol. 23, pp. 4085-4093, 2011.

[11] Q. Xiang, J. Yu, M. Jaroniec, "Graphene-based semiconductor photocatalysts," Chemical Society Reviews , vol. 41, pp. 782-796, 2012.

[12] J. G. Yu, H. G. Yu, B. Cheng, X. J. Zhao, J. C. Yu, W. K. Ho, "The effect of calcination temperature on the surface microstructure and photocatalytic activity of TiO₂ thin films prepared by liquid phase deposition," Journal of Physical Chemistry B , vol. 107, no. 50, pp. 13871-13879, 2003.

[13] J. Yu, J. Xiong, B. Cheng, S. Liu, "Fabrication and characterization of Ag-TiO₂ multiphase nanocomposite thin films with enhanced photocatalytic activity," Applied Catalysis B: Environmental , vol. 60, no. 3-4, pp. 211-221, 2005.

[14] L. Vegard, "Results of crystal analysis," Philosophical Magazine , vol. 32, pp. 505-515, 1916.

[15] B. O'Regan, M. Grätzel, "A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO₂ films," Nature , vol. 353, no. 6346, pp. 737-740, 1991.

[16] H. Tang, K. Prasad, R. Sanjinés, F. Lévy, "TiO₂ anatase thin films as gas sensors," Sensors and Actuators B , vol. 26, no. 1-3, pp. 71-75, 1995.

[17] M. A. Grela, M. A. Brusa, A. J. Colussi, "Harnessing excess photon energy in photoinduced surface electron transfer between salicylate and illuminated titanium dioxide nanoparticles," Journal of Physical Chemistry B , vol. 101, no. 51, pp. 10986-10989, 1997.

[18] Y. Wang, Y. Hao, H. Cheng, J. Ma, B. Xu, W. Li, S. Cai, "Photoelectrochemistry of transition metal-ion-doped TiO₂ nanocrystalline electrodes and higher solar cell conversion efficiency based on Zn²⁺ -doped TiO₂ electrode," Journal of Materials Science , vol. 34, no. 12, pp. 2773-2779, 1999.

[19] D. Dvoranová, V. Brezová, M. Mazúr, M. A. Malati, "Investigations of metal-doped titanium dioxide photocatalysts," Applied Catalysis B , vol. 37, no. 2, pp. 91-105, 2002.

[20] A. Fuerte, M. D. Hernández-Alonso, A. J. Maira, A. Martínez-Arias, M. Fernández-García, J. C. Conesa, J. Soria, "Visible light-activated nanosized doped-TiO₂ photocatalysts," Chemical Communications , no. 24, pp. 2718-2719, 2001.

[21] N. H. Hong, J. Sakai, W. Prellier, A. Ruyter, "Room temperature ferromagnetism in laser ablated transition-metal-doped TiO₂ thin films on various types of substrates," Journal of Physics D , vol. 38, no. 6, pp. 816-821, 2005.

[22] A. W. Xu, Y. Gao, H. Q. Liu, "The preparation, characterization, and their photocatalytic activities of rare-earth-doped TiO₂ nanoparticles," Journal of Catalysis , vol. 207, no. 2, pp. 151-157, 2002.

[23] J. Li, C. Z. Hua, "Hollowing Sn-doped TiO₂ nanospheres via Ostwald ripening," Journal of the American Chemical Society , vol. 129, no. 51, pp. 15839-15847, 2007.

[24] Y. Shen, T. Xiong, T. Li, K. Yang, "Tungsten and nitrogen co-doped TiO₂ nano-powders with strong visible light response," Applied Catalysis B: Environmental , vol. 83, no. 3-4, pp. 177-185, 2008.

[25] K. Wilke, H. D. Breuer, "The influence of transition metal doping on the physical and photocatalytic properties of titania," Journal of Photochemistry and Photobiology A , vol. 121, no. 1, pp. 49-53, 1999.

[26] M. Melemeni, D. Stamatakis, N. P. Xekoukoulotakis, D. Mantzavinos, N. Kalogerakis, "Disinfection of municipal wastewater by TiO₂ photocatalysis with UV-A, visible and solar irradiation and BDD electrolysis," Global Nest Journal , vol. 11, no. 3, pp. 357-363, 2009.

[27] D. F. Ollis, "Photocatalytic purification and remediation of contaminated air and water," C R Academic Science of Paris, Chemical , vol. 3, no. 6, pp. 405-411, 2000.

[28] A. Mills, S. Le Hunte, "An overview of semiconductor photocatalysis," Journal of Photochemistry and Photobiology A , vol. 108, no. 1, pp. 1-35, 1997.

[29] M. R. Hoffmann, S. T. Martin, W. Choi, D. W. Bahnemann, "Environmental applications of semiconductor photocatalysis," Chemical Reviews , vol. 95, no. 1, pp. 69-96, 1995.

[30] K. Kabra, R. Chaudhury, R. L. Sawney, "Treatment of hazardous organic and inorganic compounds through aqueous-phase photocatalysis: a review," Indian Engineering Chemical Research , vol. 43, no. 24, pp. 7683-7696, 2004.

[31] J. M. Herrmann, "Heterogeneous photocatalysis: state of the art and present applications," Topics in Catalysis , vol. 34, no. 1-4, pp. 49-65, 2005.

[32] G. Palmisano, V. Augugliaro, M. Pagliaro, L. Palmisano, "Photocatalysis: a promising route for 21st century organic chemistry," Chemical Communications , no. 33, pp. 3425-3437, 2007.

[33] A. P. Zhang, Y. P. Sun, "Photocatalytic killing effect of TiO₂ nanoparticles on Ls-174-t human colon carcinoma cells," World Journal of Gastroenterology , vol. 10, no. 21, pp. 3191-3193, 2004.

[34] A. Fujishima, T. N. Rao, D. A. Tryk, "Titanium dioxide photocatalysis," Journal of Photochemistry and Photobiology C , vol. 1, no. 1, pp. 1-21, 2000.

[35] C. Mccullagh, J. M. C. Robertson, D. W. Bahnemann, P. K. J. Robertson, "The application of TiO₂ photocatalysis for disinfection of water contaminated with pathogenic micro-organisms: a review," Research on Chemical Intermediates , vol. 33, no. 3-5, pp. 359-375, 2007.

[36] J. Chae, J. Lee, J. H. Jeong, M. Kang, "Hydrogen production from photo splitting of water using the Ga-incorporated TiO₂ s prepared by a solvothermal method and their characteristics," Bulletin of Korean Chemical Society , vol. 30, no. 2, pp. 302-308, 2009.

[37] C. M. Whang, J. G. Kim, E. Y. Kim, Y. H. Kim, W. I. Lee, "Effect of Co, Ga, and Nd additions on the photocatalytic properties of TiO₂ nanopowders," Glass Physics and Chemistry , vol. 31, no. 3, pp. 390-395, 2005.

[38] S. Song, C. Wang, F. Hong, Z. He, Q. Cai, J. Chen, "Gallium- and iodine-co-doped titanium dioxide for photocatalytic degradation of 2-chlorophenol in aqueous solution: Role of gallium," Applied Surface Science , vol. 257, no. 8, pp. 3427-3432, 2011.

[39] J. Zhou, Y. Zhang, X. S. Zhao, A. K. Ray, "Photodegradation of benzoic acid over metal-doped TiO₂ ," Industrial and Engineering Chemistry Research , vol. 45, no. 10, pp. 3503-3511, 2006.

[40] T. Okajima, T. Yamamoto, M. Kunisu, S. Yoshioka, I. Tanaka, N. Umesaki, "Dilute Ga dopant in TiO₂ by X-ray absorption near-edge structure," Japanese Journal of Applied Physics Part 1 , vol. 45, no. 9, pp. 7028-7031, 2006.

[41] W. Choi, A. Termin, M. R. Hoffmann, "The role of metal ion dopants in quantum-sized TiO₂ : correlation between photoreactivity and charge carrier recombination dynamics," Journal of Physical Chemistry , vol. 98, no. 51, pp. 13669-13679, 1994.

[42] B. D. Cullity Elements of X-ray Diffraction , Addison-Wesley, Reading, Mass, USA, 1978.

[43] R. Wang, N. Sakai, A. Fujishima, T. Watanabe, K. Hashimoto, "Studies of surface wettability conversion on TiO₂ single-crystal surfaces," Journal of Physical Chemistry B , vol. 103, no. 12, pp. 2188-2194, 1999.

[44] N. L. Wu, M. S. Lee, Z. J. Pon, J. Z. Hsu, "Effect of calcination atmosphere on TiO₂ photocatalysis in hydrogen production from methanol/water solution," Journal of Photochemistry and Photobiology A , vol. 163, no. 1-2, pp. 277-280, 2004.

[45] J. C. Yu, J. Yu, J. Zhao, "Enhanced photocatalytic activity of mesoporous and ordinary TiO₂ thin films by sulfuric acid treatment," Applied Catalysis B: Environmental , vol. 36, no. 1, pp. 31-43, 2002.

[46] R. V. Ghita, C. Logofatu, C. Negrila, A. S. Manea, M. Cernea, V. Ciupina, M. F. Lazarescu, "XPS study of Ti/oxidized GaAs interface," Journal of Optoelectronics and Advanced Materials , vol. 8, no. 1, pp. 31-36, 2006.

[47] R. Rao, A. M. Rao, B. Xu, J. Dong, S. Sharma, M. K. Sunkara, "Blueshifted Raman scattering and its correlation with the [110] growth direction in gallium oxide nanowires," Journal of Applied Physics , vol. 98, no. 9, 2005.

[48] P. W. Wang, Y. P. Song, X. Z. Zhang, J. Xu, D. P. Yu, "Transformation from β -Ga₂ O₃ to GaN nanowires via nitridation," Chinese Physics Letters , vol. 25, no. 3, pp. 1038-1041, 2008.

[49] J. D. Lindberg, D. G. Snyder, "Determination of the optical absorption coefficient of powdered materials whose particle size distribution and refractive indices are unknown," Applied Optics , vol. 12, no. 3, pp. 573-578, 1973.

[50] N. Barka, S. Qourzal, A. Assabbane, A. Nounah, Y. Ait-Ichou, "Factors influencing the photocatalytic degradation of Rhodamine B by TiO₂ -coated non-woven paper," Journal of Photochemistry and Photobiology A , vol. 195, no. 2-3, pp. 346-351, 2008.

[51] K. Yu, S. Yang, H. He, C. Sun, C. Gu, Y. Ju, "Visible light-driven photocatalytic degradation of rhodamine B over NaBiO₃ : pathways and mechanism," Journal of Physical Chemistry A , vol. 113, no. 37, pp. 10024-10032, 2009.

[52] Q. Xiang, J. Yu, P. K. Wong, "Quantitative characterization of hydroxyl radicals produced by various photocatalysts," Journal of Colloid and Interface Science , vol. 357, no. 1, pp. 163-167, 2011.

[53] Q. Xiang, J. Yu, M. Jaroniec, "Nitrogen and sulfur co-doped TiO₂ nanosheets with exposed {001} facets: synthesis, characterization and visible-light photocatalytic activity," Physical Chemistry Chemical Physics , vol. 13, no. 11, pp. 4853-4861, 2011.

[54] N. Guettaï, H. Ait Amar, "Photocatalytic oxidation of methyl orange in presence of titanium dioxide in aqueous suspension. Part II: kinetics study," Desalination , vol. 185, no. 1-3, pp. 439-448, 2005.

[55] X. Qin, L. Jing, G. Tian, Y. Qu, Y. Feng, "Enhanced photocatalytic activity for degrading Rhodamine B solution of commercial Degussa P25 TiO₂ and its mechanisms," Journal of Hazardous Materials , vol. 172, no. 2-3, pp. 1168-1174, 2009.

[56] J. Yu, Q. Xiang, M. Zhou, "Preparation, characterization and visible-light-driven photocatalytic activity of Fe-doped titania nanorods and first-principles study for electronic structures," Applied Catalysis B , vol. 90, no. 3-4, pp. 595-602, 2009.